Cardiology Department, S. Maugeri Fnd, IRCCS, Scientific Institute of Veruno (NO), Veruno, NO, Italy.
Eur J Nucl Med Mol Imaging. 2011 Mar;38(3):499-508. doi: 10.1007/s00259-010-1654-3. Epub 2010 Nov 11.
A new iterative reconstruction algorithm (WBR™) has been recently proposed for cardiac single photon emission computed tomography (SPECT). The WBR™ technology is designed to reduce noise, improving lesion identification without affecting the image resolution, allowing SPECT studies with reduced count statistic. This allows for either half-time (HT) or half-dose (HD) cardiac SPECT, with image quality and quantitative data comparable to standard-time (ST) or standard-dose (SD) SPECT. Few data exist on the comparison between conventional filtered backprojection (FBP) and this new algorithm in a clinical setting. The aim of this study was to compare the performance of FBP and WBR™.
Phantoms studies were performed to compare spatial resolution and contrast recovery with FBP, ordered subset expectation maximization (OSEM) and WBR™. A group of 92 patients, with different cardiac pathology, scheduled for a stress-rest SPECT were studied: 52 patients (group A) were injected with a SD of tracer and underwent both ST and HT SPECT; 40 patients (group B) were injected with a half dose of tracer and underwent ST SPECT and immediately after an additional SPECT at double time/projection (DT), to compensate for the low count statistic. A 2-day (99m)Tc-sestamibi protocol was used in all patients. SD/ST and HD/DT SPECT were reconstructed with a conventional FBP; SD/HT and HD/ST SPECT were reconstructed with WBR™. The summed stress score (SSS) and summed rest score (SRS) were calculated; the left ventricular ejection fraction (LVEF) was automatically derived.
In group A (SD), no significant differences were observed between ST FBP SPECT and HT WBR™ in SSS (11.1 and 11.7, respectively) and SRS (9.4 and 10.3, respectively, NS). LVEF on rest acquisitions was also comparable (50% on ST SPECT and 49% on HT SPECT, NS); LVEF on post-stress studies in HT SPECT (46%) was lower than ST SPECT (50%), although not statistically significant. In group B (HD), SSS (6.2 in ST and 5.3 in DT) and SRS (4.0 in ST and 3.3 in DT) were also comparable. No differences were documented between ST and DT in rest (47 and 48%, respectively) and stress (48 and 50%, respectively) LVEF.
WBR™ performance and image quality were comparable to those of conventional FBP, allowing for either HT or HD studies. The former allows for an increased patient throughput and optimization of resources. The latter modalities would allow for a significant reduction in both patients' and operators' exposure. Further studies are needed to validate the clinical use of this method.
最近提出了一种新的迭代重建算法（WBR™）用于心脏单光子发射计算机断层扫描（SPECT）。WBR™技术旨在降低噪声，提高病变识别能力，而不影响图像分辨率，允许使用减少计数统计的 SPECT 研究。这允许进行半时间（HT）或半剂量（HD）心脏 SPECT，图像质量和定量数据与标准时间（ST）或标准剂量（SD）SPECT 相当。在临床环境中，关于常规滤波反投影（FBP）和这种新算法之间的比较的数据很少。本研究的目的是比较 FBP 和 WBR™的性能。
进行了体模研究，以比较 FBP、有序子集期望最大化（OSEM）和 WBR™的空间分辨率和对比度恢复。一组 92 名患有不同心脏疾病的患者计划进行应激-静息 SPECT：52 名患者（A 组）接受 SD 示踪剂注射，并进行 ST 和 HT SPECT；40 名患者（B 组）接受半剂量示踪剂注射，并进行 ST SPECT，然后立即进行双倍时间/投影（DT）的额外 SPECT，以补偿低计数统计。所有患者均使用 2 天（99m）Tc-sestamibi 方案。SD/ST 和 HD/DT SPECT 使用常规 FBP 重建；SD/HT 和 HD/ST SPECT 使用 WBR™重建。计算总和应激评分（SSS）和总和静息评分（SRS）；自动得出左心室射血分数（LVEF）。
在 A 组（SD）中，ST FBP SPECT 和 HT WBR™ 在 SSS（分别为 11.1 和 11.7）和 SRS（分别为 9.4 和 10.3，无统计学意义）方面无显著差异。静息采集的 LVEF 也相当（ST SPECT 为 50%，HT SPECT 为 49%，无统计学意义）；HT SPECT 上的应激后研究中的 LVEF（46%）低于 ST SPECT（50%），尽管无统计学意义。在 B 组（HD）中，SSS（ST 为 6.2，DT 为 5.3）和 SRS（ST 为 4.0，DT 为 3.3）也相当。ST 和 DT 之间的静息（分别为 47%和 48%）和应激（分别为 48%和 50%）LVEF 无差异。
WBR™的性能和图像质量与传统 FBP 相当，允许进行 HT 或 HD 研究。前者允许增加患者吞吐量和优化资源利用。后者将允许显著减少患者和操作人员的暴露。需要进一步的研究来验证该方法的临床应用。
